Isozymes of Human Phosphofructokinase

Proc. Nat{. Acad. Sci. USA Vol. 77, No. 1, pp. 62-66, January 1980 Biochemistry

Isozymes of human phosphofructokinase: Identification and subunit structural characterization of a new system (hemolytic anemia/myopathy/in vitro protein hybridization/column chromatography) SHOBHANA VORA*, CAROL SEAMAN*, SUSAN DURHAM*, AND SERGIO PIOMELLI* Division of Pediatric Hematology, New York University School of Medicine, 550 First Avenue, New York, New York 10016 Communicated by Saul Krugman, July 13, 1979

ABSTRACT The existence of a five-membered isozyme The clinical effects of the enzymatic defect consisted of system for human phosphofructokinase (PFK; ATP:D-fructose- gen- 6-phosphate 1-phosphotransferase, EC 2.7.1.11) has been dem- eralized muscle weakness and compensated hemolysis. The onstrated. These multimolecular forms result from the random differential tissue involvement led to the hypothesis that the polymerization of two distinct subunits, M (muscle type) and erythrocyte isozyme is composed of two types of subunits, one L (liver type), to form all possible tetrameters-i.e., M4, M3L, of which is the sole subunit present in muscle PFK (9, 10). The M2L4, ML3, and L4. Partially purified muscle and liver PFKs proposed structural heterogeneity of erythrocyte PFK protein were hybridized by dissociation at low pH and then recombi- was nation at neutrality. Three hybrid species were generated in supported by immunochemical neutralization experiments addition to the two parental isozymes, to yield an entire five- (11, 12). Karadsheh et al. (13) and Kaur and Layzer (14) have membered set. The various species could be consistently and recently presented data to support the suggested hybrid reproducibly separated from one another by DEAE-Sephadex structure for erythrocyte PFK. chromatography at pH 8.0 with a concave elution gradient of The present study addressed the following three basic salt. Under similar experimental conditions, erythrocyte PFK questions: Could PFK be a from hemolysates was also resolved into five species chroma- (i) erythrocyte mixture of all five tographiclly indistinguishable from those produced in the theoretically possible tetramers composed of two distinct sub- above experiment. Immunological and kinetic studies of the units, rather than a single hybrid species? (ii) What is the tissue isozymes provided corroborative evidence to support the pro- of origin of the nonmuscle subunit identified as an "E" or "R" posed subunit structures. Erythrocyte PFK was found to have subunit (13, 14) by other workers? (ii) What are the relative kinetic properties intermediate between those of muscle and proportions of these two subunits in the liver PFK and was neutralized only 50% by an antiserum erythrocytes? against muscle PFK that completely neutralized muscle PFK. Preliminary results of these studies have been presented These data demonstrate that muscle and liver PFKs are distinct (15). homotetramers-i.e., M4 and L4, respectively-whereas erythrocyte PFK is a heterogeneous mixture of all five isozymes. MATERIALS AND METHODS The structural heterogeneity of erythrocyte PFK provides a molecular genetic basis for the differential organ involvement Chemicals and Reagents. Adenosine nucleotides, NAD, observed in some inherited PFK deficiency states in which NADH, Fru-6-P, Fru-1,-P2, dithiothreitol, glycylglycine, and myopathy or hemolysis or both can occur. sodium glycerophosphate were purchased from Sigma; auxiliary enzymes were from Boehringer Mannheim; Freund's Like most allosteric enzymes, phosphofructokinase (PFK; complete and incomplete adjuvants were from Difco; DEAE- ATP:D-fructose-6-phosphate 1-phosphotransferase, EC cellulose (DE-52) was from Whatman; DEAE-Sephadex A-25 2.7.1.11), a key enzyme of glycolysis, is a multimeric protein, was from Pharmacia; agarose-hexane-ATP (AGATP), type 2, the smallest active oligomer being a tetramer. The enzyme from was from P-L Biochemicals; Ionagar was from Colab Labs a large number of plant, bacterial, nonvertebxate, and verte- (Chicago Heights, IL); Nonidet was from Particle Data (Chi- brate species has been purified and studied. The physico- cago, IL). Human muscle and liver specimens, obtained at chemical and kinetic properties of PFKs from various mam- surgery and autopsy, respectively, were stored frozen at -200C malian tissues have been investigated, and in a few instances, until used for extraction. Venous blood was obtained from multimolecular forms have been noted (1). At least four forms healthy volunteers. of PFK have been reported in tumors and normal tissues of rats Assays of PFK Activity. Assays were done with a Gilford on the basis of ion-exchange chromatography elution profiles model 2400 spectrophotometer at 260C. The reaction mixture and of antibody precipitation tests (2, 3). The extensive studies (1.0 ml) contained 50 mM tris(hydroxymethyl)methylglycine by Tsai and Kemp (4-7) have demonstrated the presence of a at pH 8.4, 5 mM MgC12, 0.15 mM NADH, 0.1 mM di- multilocus isozyme system for PFK in the rabbit. In this species, thiothreitol, 0.25 mM ATP, 2.4 mM Fru-6-P, 0.18 unit of al- three separate genetic loci code for a structurally different dolase, 0.6 unit of triose phosphate isomerase, and 0.1 unit of polypeptide subunit of PFK; either one or two of these poly- glycerophosphate dehydrogenase. The reaction was started by peptides are expressed in each tissue, resulting in the formation addition of 0.1 ml of the enzyme preparation to 0.9 ml of the of either single (homotetrarneric) or multiple (homo- and het- assay mixture. The decrease in absorbance at 340 nm was re- erotetramertic) isozymes. corded for the next 15-20 min against a blank from which PFK The existence of an isozyme system for PFK in man was first or Fru-6-P was omitted. One unit of enzyme is defined as that suggested in 1965 by the observation of a recessively inherited amount of enzyme that converts 1 Mimol of Fru-6-P to Fru- muscle disease associated with PFK deficiency in three siblings 1,6-P2 in 1 min in the above system. of a Japanese family (8). In these individuals, PFK activity was entirely absent in muscle and was half normal in erythrocytes. Abbreviations: PFK, ATP:D-fructose-6-phosphate I-phosphotransferase, EC 2.7.1.11; Ko.5, concentration of the substrate required to The publication costs of thi$ article were defrayed in part by page achieve half-maximal velocity; M, muscle subunit; L, liver subunit. charge payment. This article must therefore be hereby marked "ad- * Present Address: College of Physicians and Surgeons, Columbia vertisement" in accordance with 18 U. S. C. §1734 solely to indicate University, Division of Pediatric Hematology, Department of Pe- this fact. diatrics, 3959 Broadway, New York, NY 10032. 62 Downloaded by guest on September 28, 2021 Biochemistry: Vora et al. Proc. Natl. Acad. Sci. USA 77 (1980) 63 The kinetics of the PFK isozymes were studied at pH 7.2 in Dissociation and Reassociation Experiments with PFKs 50 mM Tris-HCI/1 mM EDTA/0.2 mM dithiothreitol. The from Muscle and Liver. These were performed according to concentrations of Mg9l2, NADH, and auxiliary enzymes were the procedure described by Tsai and Kemp (4) with some the same as described above. Ammonium sulfate was removed modifications as described below. Equal amounts (in units) of from the auxiliary enzymes by dialysis against Tris.HCI buffer partially purified muscle and liver PFKs were dissociated at 4VC for 2 hr. All reactions were started by the addition of separately by dissolving them in small volumes of 0.025 M and Fru-6-P. To determine the apparent concentration of Fru-6-P 0.1 M glycylglycine/fl-glycerophosphate buffers (pH 5.2), required to achieve half-maximal velocity (Ko.5) values, ATP respectively, each containing 1 mM EDTA and 0.25 mM di- was kept constant (0.1 mM) and Fru-6-P concentration was thiothreitol. Muscle and liver PFK solutions were incubated at varied from 0.025 to 1.6 mM. Vnx was measured by activation 26'C for 2 and 10 min, respectively. The duration of exposure with 0.2 mM AMP. The KO.5 Fru-6-P value for each isozyme to low pH and the strength of the buffer for each enzyme were was calculated from the Hill plots. determined by trial and error, so as to achieve >95% loss of Preparation of PFKs from Muscle, Liver, and Erythro- activity, suggesting maximal dissociation. An aliquot of each cytes. Muscle PFK was partially purified according to the solution was removed for assay, and equal volumes (one-half method of Kemp (16) except that the crystallization steps were of the original volumes) were then mixed. Reassociation of both omitted. For the immunization of rabbits, this preparation was the mixture and the remaining original solutions was brought subjected to DEAE-cellulose chromatography (17) followed about by adjusting the pH to 7.2 with 0.5 M Tris phosphate at by affinity chromatography (18). The final preparation yielded pH 8.5. The solutions were incubated at 26°C for another 10 a single band on sodium dodecyl sulfate/polyacrylamide gel min and assayed again. The reassociation of isozymes was in- electrophoresis. Protein determination was according to Lowry dicated by the recovery of enzymatic activities. Independently et al. (19). Partial purification of liver PFK was carried out reassociated muscle and liver PFKs served as controls. In a series according to the method of Brock (20), and that of erythrocyte of similar experiments, muscle and liver isozymes were hy- PFK was according to the procedure described by Wenzel et bridized in varying proportions (3:1 and 1:3). al. (21) except that, in both cases, the final gel filtration steps Dissociation and Reassociation of Erythrocyte PFX. were omitted. Partially purified erythrocyte PFK was subjected to dissociating Chromatographic Separation of PFK Isozymes. All the and reassociating conditions as described above for liver operations were carried out at 4VC. The columns were loaded PFK. with 0.1-0.15 unit of the enzyme preparation. For linear gra- Preparation of Anti-Human Muscle PFK Antibody. For dient elution, a DEAE-Sephadex A-25 column (1.7 X 30 cm) primary immunization, each of the five New Zealand White was equilibrated in 0.1 M Tris phosphate, pH 8.0/0.2 mM female rabbits received 250 ,g of purified muscle PFK sus- EDTA/0.2 mM AMP/0.7 mM dithiothreitol. A 100-ml linear pended in 0.5 ml of potassium phosphate, pH 8.0/0.2 mM gradient of NaCl from 0 to 0.8 M prepared in the equilibrating EDTA/10 mM (NH4)2S04/0.1 mM ATP/0.1 mM dithiothreitol buffer was used for elution; 100 fractions were collected. For and emulsified with 0.5 ml of Freund's complete adjuvant; 0.25 concave gradient elution, the gel was equilibrated in the buffer ml was injected into each footpad (1 ml per rabbit). Two booster mentioned above, to which 0.025 M NaCI was added. Elution doses were given subcutaneously 2-3 weeks apart, containing was done with a 300-ml concave gradient of NaCI from 0.025 the same amounts of antigen but in Freund's incomplete ad- to 0.525 M in the same buffer; 100 fractions were collected. juvant. Test bleedings were done 4-6 weeks after the primary

M2 L2 |A B

- 500 r l~~~~~~~~~~~~~~~~i 0) s C D Eb. 0 Yi E ,' FIG. 1. Chromato- graphic separation of PFKs from muscle, liver, and erythrocytes. (A and B) Resolution of a mixture of muscle (M4) and liver (L4) isozymes with the linear and concave elution gradients, respectively. (C and D) Resolution of erythrocyte PFK with the linear and concave elution gradi- Fraction ents, respectively. Downloaded by guest on September 28, 2021 64 Biochemistry: Vora et al. Proc. Natl. Acad. Sci. USA 77 (1980)

li I-

FIG. 2. Chromato- graphic resolution ofdisso- ciated and reassociated PFK isozymes. (A) Hy- bridized solution of muscle and liver PFKs in 1:1 ratio. (B) Mixture of control solutions of muscle and liver from A. (C) Dissociated and reassociated erythrocyte isozymes. (D) Hybridiza- tion of muscle and liver Fraction PFKs in 3:1 ratio.

immunizations and 1 week after the booster doses, and two solved into a broad major fraction with its peak in a position just animals that showed high antibody titer were bled further. overlapping the right descending limb of muscle PFK and a Control sera were obtained by bleeding the animals prior to minor fraction with its peak in the position of the liver isozyme immunizations. (Fig. 1C). Immunologic Studies. Precipitation of enzyme with anti- Muscle and liver preparations were recovered quantitatively serum was carried out as follows. Partially purified PFK (90 + 10%) when either crude tissue homogenates or partially preparations were diluted with 50 mM Tris.HCI, pH 8.0/1 mM purified preparations were chromatographed. Crude eryth- EDTA/1 mM ATP/1 mM E-aminocaproic acid/i mM di- rocyte hemolysates yielded quantitative recovery but partially thiothreitol/0.2% bovine serum albumin in a final concentration purified ones did not. The poor recoveries (approximately 30%) of 0.2-0.6 unit/ml. In a series of test tubes, 0.3 ml of diluted of the latter may be explained by the tendency of human enzyme was mixed with 0.1 ml of antibody diluted serially with erythrocyte PFK to undergo self-aggregation to high molecular control rabbit serum. The mixtures of enzymes and diluted weight species, with a concomitant decline in specific activity, antisera were incubated for 30 min at 370C and then centri- at high protein concentrations (22). fuged for 20 min at 15,000 X g. The enzyme activity in the Chromatographic Separation of Native PFKs on Concave supernatant was assayed at pH 8.4. Ouchterlony analysis was Elution Gradients. The broad major peak of erythrocyte PFK carried out according to the method described by Tsai and was resolved into four peaks when the early phase of the linear Kemp (5) with minor modifications. Double-diffusion plates eluting gradient was expanded by using a concave gradient contained 0.8% agar in 50 mM Tris phosphate, pH 8.0/2 mM (Fig. 1D). The first and last peaks eluted in the same positions EDTA/0.1 mM ATP/0.1 mM (NH4)2SO4/0.1 mM di- as muscle and liver isozymes, respectively (Fig. 1B). thiothreitol/1% Nonidet P-40. The gel was prepared by heating The muscle- and liver-type species of native erythrocyte PFK 1 vol of 1% agar solution to boiling and then rapidly mixing in could be quantitated with some degree of accuracy and were 0.25 vol of buffer containing all the reagents listed above but found to comprise 1% and 6-11%, respectively, of the total at 5 times the concentrations. The agar was immediately poured enzyme. However, the hybrids could not be quantitated, due into petri dishes. Wells were punched and plates were used to the overlap of their peaks. There were minor variations in within 24 hr. The wells (15 ,l) were filled and diffusion was the isozymic profile of erythrocyte PFK from one individual carried out at 37°C for 24 hr. to another but in any given individual the profile was found to be consistent and reproducible. RESULTS Dissociation and Reassociation of PFKs from Muscle and Chromatographic Separation of Native PFKs on Linear Liver. Muscle and liver PFKs lost more than 95% of their ac- Elution Gradients. Muscle and liver PFK each eluted as a tivities after standing at pH 5.2 for variable lengths of time. On single, sharp, symmetrical peak widely separated from the other reassociation at pH 7.2, the muscle, liver, and hybridized so- (Fig. 1A). This technique, however, could not resolve muscle lutions regained approximately 40, 60, and 50% of their ac- and erythrocyte PFKs distinctly from each other. Under tivities, respectively. The loss of enzymatic activity is probably identical experimental conditions, erythrocyte PFK was re- due to irreversible denaturation of the subunits at low pH. Five Downloaded by guest on September 28, 2021 Biochemistry: Vora et al. Proc. Natl. Acad. Sci. USA 77 (1980) 65 isozymes were produced when hybridization was performed (11, 12). Muscle and erythrocyte enzymes were also shown to with equal amounts of muscle (M) and liver (L) enzymes (Fig. be separable by DEAE-cellulose chromatography and a step- 2A). The three hybrid species produced in vitro, as well as the wise elution gradient of salt (12). Recent studies have shown that two parental isozymes, were eluted at the same positions in the human erythrocyte PFK is composed of two nonidentical gradient as the naturally occurring isozymes from erythrocytes. subunits, of 85,000 and 80,000 daltons, respectively. The The selective loss of M subunits was reflected in the relative 85,000-dalton subunit is designated "muscle type" because it proportions of these five species. The chromatogram resembled is found to be identical to the sole subunit present in muscle PFK an ascending cascade rather than a symmetrical binomial dis- on the basis of their tryptic peptide maps (13). The 80,000- tribution. The control solutions, which were individually dalton subunit is assumed to be specific for erythrocytes and reassociated and mixed just prior to chromatography, showed is designated as "E (erythrocyte)" or "R (red cell)" subunit. The the presence of two parental isozymes and a third very small relative proportions of these two have been variably reported peak in the position of ML3 (Fig. 2B). This small peak, also to be 1:2.9 (13) and 1.2 or less (14). However, none of these obtained when the liver preparation alone was dissociated or studies identified the nature of the nonmuscle subunit or ad- reassociated, probably was derived from blood contamination dressed the question of why only a single heterotetrameric or was indigenous to nonparenchymal cells (e.g., Kupffer cells) molecule would be assembled in erythrocytes, rather than the and fibrous stroma. entire five-membered set, if two distinct subunits were In a series of hybridization experiments, when the M:L ratio present. was increased from 1 to 1.5, a symmetrical binomial pattern Our data demonstrate that the nonmuscle subunit of eryth- was obtained (data not shown). Fig. 2D illustrates the de- rocyte PFK is similar to the subunit present in liver PFK, the scending cascade-like chromatogram when the M:L ratio was L type. With in vitro molecular hybridization experiments, further increased to 3:1. human muscle and liver enzymes have been identified as dis- Dissociation and Reassociation of Erythrocyte Isozymes. tinct homotetramers (M4 and L4), whereas erythrocyte PFK Partially purified erythrocyte PFK lost more than 95% of its has been shown to be a heterogeneous mixture of five isozymes activity at low pH and regained approximately 50% at neu- (M4, M3L, M2L2, ML3, and L4), with the two subunits being trality. The chromatographic profile of dissociated and reas- present in approximately equal proportions. sociated erythrocyte PFK was nearly identical to that produced We initially chose to determine whether human skeletal by the hybridization of muscle and liver PFKs in 1:1 proportion muscle and liver PFKs were truly distinct homotetrameric (Fig. 2C). isozymes or not, as they are in the case of the rabbit (4,33, 34). Immunologic Studies. In double-diffusion studies, muscle Although indistinguishable by electrophoresis, these two were and erythrocyte PFK each yielded a single precipitin line, and readily resolved by DEAE-Sephadex chromatography. The there was a reaction of identity; liver isozyme failed to react resolution of erythrocyte PFK into two peaks suggested that the altogether. An excess of antibody precipitated >90% of muscle PFK and minor peak may be similar to the liver isozyme and the major approximately 50% of normal erythrocyte PFK but one may consist of either one or many hybrid isozymes, com- <5% of liver PFK. posed of both muscle and liver type subunits. Kinetic Studies. All three isozymes exhibited sigmoidal ki- When the broad peak of erythrocyte PFK was further re- netics in the absence of an activator, with different affinities solved by using a concave gradient, it separated into four peaks. for Fru-6-P. The K0.5 Fru-6-P value of the erythrocyte PFK The first peak eluted in the position of muscle PFK, three others (0.46 mM) was double that of the liver isozyme (0.24 mM) and appeared in positions intermediate between muscle and liver half of that of the muscle isozyme (0.89 mM). Upon addition isozymes, and the last peak was in the position of liver PFK. This of AMP, the sigmoidal curves shifted to hyperbolic ones, finding appeared consistent with the presence in the erythro- showing classical substrate-saturation kinetics. The K0.5 values cyte of five different tetrameric isozymes, resulting from all in the presence of AMP were 0.11 mM for muscle PFK and 0.03 the possible combinations of two distinct subunits (M and L): mM for liver PFK. The K0.5 Fru-6-P of erythrocyte PFK (0.06 two parental homotetramers, M4 and L4, and three hybrids, mM) was again intermediate between those of the other two M3L, M2L2, and ML3. enzymes. The hypothesis that these five chromatographically distin- DISCUSSION guishable isozymes were members of a single isozyme family was tested by the hybridization experiments, which exploit the Considerable metabolic diversity is observed in the regulation general property of mammalian PFK to undergo reversible of carbohydrate metabolism in various tissues of a given species dissociation (1, 35). The dissociation of tetramers into dimers (23). Control of glycolysis is usually explained in terms of the and monomers can be brought about by lowering the pH of the allosteric properties of three enzymes: PFK, hexokinase, and enzyme solution to below 5.8 and is accompanied by a loss of pyruvate kinase. Generally, the PFK step is considered to be enzyme activity to 2-5% of the original. A subsequent increase the major rate-limiting reaction (24, 25). The intracellular in pH to above 7.0 leads to reassociation into tetramers with catalytic activities of the enzymes are profoundly influenced partial recovery of enzyme activity (33). by various metabolites. Thus, diverse metabolic regulation in Dissociation and reassociation of human muscle and liver various tissues can be achieved by differences in effector con- PFKs yielded three additional hybrid species, as well as the centrations or by the presence of tissue-specific isozymes with original homotetramers. In the control experiment, when different physicochemical, kinetic, and regulatory properties. muscle and liver PFKs were mixed after independent disso- Our studies indicate that both of the above mechanisms are ciation and reassociation, the two original homotetramers were operative with respect to human PFK. obtained, along with a small peak in the ML3 position. This ML3 The original description of inherited muscle PFK defi- peak was due to muscle subunits contaminating the liver ciency-glycogen storage disease, type VIII (26)-in 1965 by preparation. Tarui et al. (8) prompted various investigators to examine the These results were consistent with the concept that muscle physicochemical and kinetic properties as well as the genetic and liver PFKs are different proteins with homotetrameric control of this enzyme (27-32). The hypothesis of a common structures. muscle-type subunit in both muscle and erythrocyte PFKs was The random recombination of equal numbers of two distinct initially suggested by immunochemical neutralization studies subunits (M and L) is expected to yield five isozymes in a 1:4: Downloaded by guest on September 28, 2021 66 Biochemistry: Vora et al. Proc. Natl. Acad. Sci. USA 77 (1980)

6:4:1 ratio, as predicted by the binomial distribution, provided 1. Bloxham, D. P. & Lardy, H. A. (1973) in The Enzymes, ed. that both subunits exhibit similar intra- and interspecies af- Boyer, P. D. (Academic, New York), pp. 239-278. finities in vitro. The deviation from a binomial distribution of 2. Tanaka, T., An, T. & Sakaue, Y. (1971) J. Biochem. 69, 609- 612. five peaks in our study reflects the excessive loss of M subunits 3. Taylor, C. B. & Bew, M. (1970) Biochem. J. 119,797-799. during low pH treatment. In fact, during hybridization, the M:L 4. Tsai, M. Y. & Kemp, R. G. (1972) Arch. Biochem. Biophys. 150, ratio was increased to 1.5:1, and a symmetrical distribution was 407-411. evident. A further excess of M or L subunits yielded isozyme 5. Tsai, M. Y. & Kemp, R. G. (1973) J. Biol. Chem. 248, 785- profiles that were highly skewed toward M- or L-containing 792. 6. Tsai, M. Y. & Kemp, R. G. (1974) J. Biol. Chem. 249, 6590- species, respectively. These data indicated that in vitro, inter- 6596. species and intraspecies affinities of each subunit were 7. Tsai, M. Y., Gonzalez, F. & Kemp, R. G. (1975) in Isozymes, ed. equal. Markert, C. L. (Academic, New York), Vol. 2, pp. 819-835. Partially purified erythrocyte PFK, after dissociation and 8. Tarui, S., Okuno, G., Ikura, Y., Tanaka, T., Suda, M. & Nishikawa, reassociation, yielded an isozymic profile nearly identical to M. (1965) Biochem. Biophys. Res. Commun. 19,517-523. that produced by muscle and liver PFKs hybridized in equal 9. Layzer, R. B., Rowland, L. P. & Ranney, H. M. (1967) Arch. Neurol. 17, 512-523. proportions. This finding suggested that, in native erythrocyte 10. Tarui, S. (1967) Proceedings of the 17th General Assembly of the PFK, the ratio of M to L subunits cannot be very different from Japan Medical Congress, ed. Kanda, Z. (Japanese Medical Lit- unity. This conclusion is also supported by the clinical obser- erature Publishing Association, Nagoya, Japan) 3, 512-514. vation that patients genetically lacking M subunits exhibit 11. Layzer, R. B., Rowland, L. P. & Bank, W. J. (1969) J. Biol. Chem. half-normal levels of erythrocyte PFK (8, 9, 36). 244,3823-3831. 12. Tarui, S., Kono, N., Nasu, T. & Nishikawa, M. (1969) Biochem. Our immunological and kinetic studies provide further Biophys. Res. Commun. 34,77-83. support to the proposed subunit structures of PFK isozymes. 13. Karadsheh, N. S., Uyeda, K. & Oliver, R. M. (1977) J. Biol. Chem. An excess of anti-human muscle PFK serum precipitated >90% 252,3515-3524. of muscle PFK and approximately 50% of erythrocyte PFK, as 14. Kaur, J. & Layzer, R. B. (1977) Biochem. Genet. 15, 1133- shown by other investigators (11, 12). However, our experi- 1142. 15. Vora, S. & Piomelli, S. (1977) Blood 50,87 (abstr.). ments indicated that this antiserum failed to precipitate liver 16. Kemp, R. G. (1975) Methods Enzymol. 42,71-77. enzyme. In Ouchterlony analysis, muscle and liver PFK each 17. Ling, K. H., Paetkau, V., Marens, F. & Lardy, H. A. (1966) yielded a single precipitin line which showed a reaction of Methods Enzymol. 9,425-429. identity, but liver PFK failed to react. The absence of cross- 18. Ramadoss, C. S., Luby, L. J. & Uyeda, K. (1976) Arch. Biochem. reactivity between muscle and liver PFK suggested the presence Biophys. 175,487-494. 19. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. of large structural differences between M and L subunits. This (1951) J. Biol. Chem. 193,265-275. is in agreement with the reported differences in the tryptic 20. Brock, D. J. H. (1969) Biochem. J. 113,235-242. peptide maps and amino acid compositions between the two 21. Wenzel, K. W., Gauer, J., Zimmermann, G. & Hoffmann, E. subunits of erythrocyte PFK (13) and supports our hypothesis (1972) FEBS Lett. 19,281-284. that the nonmuscle subunit of erythrocyte PFK is similar to the 22. Zimmermann, G., Wenzel, K. W., Gauer, J. & Hoffmann, E. liver isozyme. The erythrocyte isozymes as a group exhibited (1973) Eur. J. Biochem. 40,501-505. 23. Scrutton, M. C. & Utter, M. F. (1968) Annu. Rev. Biochem. 37, a K0.5 Fru-6-P value (in the presence or absence of AMP) in- 249-302. termediate between those of muscle and liver PFKs. 24. Minakami, S. & Yoshikawa, H. (1966) J. Biochem. 59, 139- From the evidence presented, we conclude that human 144. muscle and liver PFKs are distinct homotetramers-i.e., M4 and 25. Rose, I. A. & Warms, J. V. B. (1966) J. Biol. Chem. 241,4848- 4854. L4-distinguishable by their physicochemical, kinetic, and 26. Brown, B.I. & Brown, D. H. (1968) in Carbohydrate Metabolism immunological properties; erythrocyte PFK, until now con- and Its Disorders, eds. Dickens, F., Randle, P. J. & Whelan, W. sidered a single hybrid isozyme, is a mixture of two homotet- J. (Academic, New York), Vol. 2, pp. 123-150. ramers and three hybrid isozymes. Our studies are consistent 27. Layzer, R. B. & Conway, M. M. (1970) Biochem. Biophys. Res. with the existence of two structural loci coding for M and L Commun. 40,1259-1265. 28. Lee, L. M. Y. (1972) Arch. Biochem. Biophys. 148,607-613. subunits, which give rise to a five-membered isozyme system. 29. Wenzel, K. W., Zimmermann, G., Gauer, J., Diezel, W., Liebe, In accordance with the recommendation of the International S. T. & Hoffmann, E. (1972) FEBS Lett. 19,285-289. Union of Biochemistry Enzyme Commission (37), we propose 30. Staal, G. E. J., Koster, J. F., Banziger, C. J. M. & Van Milligen- that liver isozyme be referred to as PFK-1, the muscle isozyme Boersma, L. (1972) Biochim. Biophys. Acta 276, 113-123. as PFK-5, and the hybrids as PFK-2, -3, and -4. 31. Layzer, R. B. & Rasmussen, J. (1974) Arch. Neurol. 31, 411- The genetic heterogeneity of erythrocyte PFK may explain 417. 32. Hoffmann, E., Kurganov, B. I., Schellenberger, W., Schulz, J., the reduction in the erythrocyte enzyme activity by approxi- Sparmann, G., Wenzel, K. W. & Zimmermann, G. (1975) Adv. mately 50% observed in inherited PFK deficiency associated Enzyme Regul. 13,247-277. with myopathy. If this syndrome is due to complete lack of the 33. Paetkau, V. & Lardy, H. A. (1967) J. Biol. Chem. 242, 2035- muscle-type PFK, the residual erythrocyte enzyme is expected 2042. 34. Coffee, C. J., Aaronson, R. P. & Frieden, C. (1973) J. Biol. Chem. to be liver-type. We have recently demonstrated the exclusive 248, 1381-1387. presence of the liver-type isozyme in the erythrocytes of such 35. Mansour, T. E. (1965) J. Biol. Chem. 240,2165-2172. an individual (36). It can be speculated that the heterogeneous 36. Vora, S., Corash, L. & Piomelli, S. (1978) Blood 52, 105 group of hemolytic syndromes associated with partial eryth- (abstr.). rocyte PFK deficiency without myopathy (38-41) may be 37. IUPAC-IUB Commission on Biochemical Nomenclature (1971) secondary to the total absence of L subunits or qualitative de- Biochemistry 10,4825-4826. 38. Waterbury, L. & Frenkel, E. P. (1972) Blood 39,415-425. fects of M or L subunits. 39. Miwa, S., Sato, T., Murao, H., Kozuru, M. & Ibayashi, H. (1972) Acta Haemotol. Jpn. 35,113-118. We thank Dr. E. Pearlstein for his help in the preparation of the 40. Boulard, M. R. & Meienhoffer, M. C. (1974) N. Engl. J. Med. 291, antibody and Mr. S. Vasta for his help in the preparation of the man- 978-979. uscript. This work was supported in part by National Institutes of 41. Kahn, A., Etiemble, J., Meienhoffer, M. C. & Bovin, P. (1975) Health Grants AM-09274-14 and AM-24074-01. Glin. Chim. Acta 61,415-419. Downloaded by guest on September 28, 2021